Voice Recognition and Mobility in the Legal Industry

47
La Salle University La Salle University Digital Commons Mathematics and Computer Science Capstones Mathematics and Computer Science, Department of Spring 5-18-2015 Voice Recognition and Mobility in the Legal Industry David Solomon La Sal, [email protected] Brian Pillar La Salle University, [email protected] Follow this and additional works at: hp://digitalcommons.lasalle.edu/mathcompcapstones Part of the Graphics and Human Computer Interfaces Commons is esis is brought to you for free and open access by the Mathematics and Computer Science, Department of at La Salle University Digital Commons. It has been accepted for inclusion in Mathematics and Computer Science Capstones by an authorized administrator of La Salle University Digital Commons. For more information, please contact [email protected]. Recommended Citation Solomon, David and Pillar, Brian, "Voice Recognition and Mobility in the Legal Industry" (2015). Mathematics and Computer Science Capstones. 22. hp://digitalcommons.lasalle.edu/mathcompcapstones/22

Transcript of Voice Recognition and Mobility in the Legal Industry

Page 1: Voice Recognition and Mobility in the Legal Industry

La Salle UniversityLa Salle University Digital Commons

Mathematics and Computer Science Capstones Mathematics and Computer Science, Departmentof

Spring 5-18-2015

Voice Recognition and Mobility in the LegalIndustryDavid SolomonLa Sal, [email protected]

Brian PillarLa Salle University, [email protected]

Follow this and additional works at: http://digitalcommons.lasalle.edu/mathcompcapstones

Part of the Graphics and Human Computer Interfaces Commons

This Thesis is brought to you for free and open access by the Mathematics and Computer Science, Department of at La Salle University DigitalCommons. It has been accepted for inclusion in Mathematics and Computer Science Capstones by an authorized administrator of La Salle UniversityDigital Commons. For more information, please contact [email protected].

Recommended CitationSolomon, David and Pillar, Brian, "Voice Recognition and Mobility in the Legal Industry" (2015). Mathematics and Computer ScienceCapstones. 22.http://digitalcommons.lasalle.edu/mathcompcapstones/22

Page 2: Voice Recognition and Mobility in the Legal Industry

Voice Recognition and Mobility in the Legal Industry

La Salle University

INL 880 Capstone

Brian Pillar, David Solomon

Page 3: Voice Recognition and Mobility in the Legal Industry

Table of Contents

Voice Recognition and Mobility in the Legal Industry.................................................................. 1

Dragon Naturally Speaking vs. Windows Voice Recognition Analysis ........................................ 1

Accuracy ................................................................................................................................ 2

Common Mistakes in Both Programs ..................................................................................... 2

Test Results ........................................................................................................................... 3

Document Management Systems and Enterprise Content Management ................................... 4

Document Management System Features .............................................................................. 4

Document Management System Tools ...................................................................................... 5

Security .................................................................................................................................. 5

Portable Document Format ..................................................................................................... 5

Optical Character Recognition ................................................................................................ 5

Metadata and Full Text Search ............................................................................................... 5

Types of Formatted Documents.............................................................................................. 6

Merge Document Example ..................................................................................................... 6

Document Merge Steps ............................................................................................................. 9

Document Merge Process .....................................................................................................10

Generating Content ...................................................................................................................16

Current Digital Dictation System Process ..............................................................................16

Dragon on the Desktop ..........................................................................................................18

Dragon via Mobile .................................................................................................................19

Analysis of Dictation Workflow ..................................................................................................21

Dragon on the Desktop ..........................................................................................................21

Measuring the Success of Dragon on the Desktop ................................................................23

Measuring Return on Investment of Dragon on the Desktop ..................................................23

Dragon via Mobile with Voice Recognition on the Server Side ...............................................23

Big Hand Costs .....................................................................................................................25

Big Hand Vs Dragon Cost Breakdown ...................................................................................26

Dragon Naturally Speaking Implementation ..............................................................................26

Hardware Requirements and Cost .........................................................................................26

Installation and Training ........................................................................................................27

Voice Recognition vs Manual Transcription Pathology Lab Study .............................................29

Conclusion ................................................................................................................................30

Page 4: Voice Recognition and Mobility in the Legal Industry

Lessons Learned ...................................................................................................................31

References ...............................................................................................................................32

Appendix 1: Narrative................................................................................................................34

Dragon Naturally Speaking (Unedited) ..................................................................................34

Windows Voice Recognition ..................................................................................................36

The Original Script (Edited By Dragon Naturally Speaking) ...................................................38

Appendix 2: Initial Case Briefing ...............................................................................................41

Initial Briefing Legal Document Dictation from a Template .....................................................41

Reflections after Dictating Initial Briefing Analysis .................................................................42

Page 5: Voice Recognition and Mobility in the Legal Industry

Table of Figures

Figure 1 – Interrogatory Template ................................................................................................................ 7

Figure 2 - Merge Codes ................................................................................................................................. 8

Figure 3 - Document Merge Steps ................................................................................................................ 9

Figure 4 – Launch New Merge Document .................................................................................................. 10

Figure 5 - Template Search and Launch ...................................................................................................... 11

Figure 6 - Select Parties ............................................................................................................................... 12

Figure 7 – Merge Document Summary and Review ................................................................................... 13

Figure 8 - Additional Non-Merge Document Information .......................................................................... 14

Figure 9 - Final Document ........................................................................................................................... 15

Figure 10 - Digital Dictation Process ........................................................................................................... 17

Figure 11 - Attorney Using Dragon on the Desktop .................................................................................... 18

Figure 12 - Attorney Using Dragon on the Server ....................................................................................... 19

Figure 13 - Screenshot of Dictation Playback with Voice Recognition ....................................................... 20

Figure 14 - Document Approval .................................................................................................................. 21

Figure 15 - Dragon Review .......................................................................................................................... 25

Figure 16 - Dragon Sidebar ......................................................................................................................... 28

Page 6: Voice Recognition and Mobility in the Legal Industry

1

Voice Recognition and Mobility in the Legal Industry

In a typical legal work environment, attorneys work with their staff to generate and send case related

legal documents and communications. Traditionally the attorney will dictate to a device capable of

recording audio and the legal assistant will transcribe the audio directly from the source. In the early

days of recorded dictation audio was recorded and saved to analog tape. Once the technology became

available, dictation was saved digitally to flash memory and transmitted to hard disk for playback by the

legal assistant. It has been this way for years, and due to advances in voice recognition technology and

computer processing there are alternative options to the traditional dictation/transcription process.

The focus of this paper is to examine the traditional process of dictation/transcription and how it

compares to the process of using voice recognition software. Analysis of each process as well as an

evaluation of voice recognition software will be performed. The document generation process will also

be examined as it relates to transcription and creating a document, regardless of the content. The most

efficient solution which benefits a small to medium size law firm will be recommended. According to

Understanding How Law Offices Do Business, a small law firm has between one and ten lawyers and a

mid-size law firm has up to 50 lawyers. These firms are the target audience.

The goal of this paper is to determine if the use of voice recognition software can help an attorney and

their staff be more efficient, and if so, which voice recognition software and methods work the best.

Tests will be performed analyzing both Dragon Naturally Speaking 12 Professional and Windows 7 voice

recognition software on the desktop. The software with the higher accuracy rate based on our tests will

be used to evaluate voice recognition processes throughout this paper.

Dragon Naturally Speaking vs. Windows Voice Recognition

Analysis

To test the accuracy and usage of both Dragon Naturally Speaking and Windows built in voice

recognition, two tests were conducted. The first consisted of reading a passage of text of 1,642 words to

get a feel for how both programs handled speech recognition without any document formatting or style.

Please refer to Appendix 1 for full scope of the document. The second test consisted of dictating an

example of a case analysis for an insurance adjuster, a common type of document in the area of

Workers’ Compensation. Appendix 2 reflects the template that was used to compose the document

through voice dictation in Dragon Naturally Speaking and Windows Voice Recognition.

Both Dragon Naturally Speaking Premium Version 12 and Windows 7 Voice Recognition programs have a

set up process which checks the microphone levels by reading passages and analyzing the voice. Both

program setups require the user to read a passage to ensure the microphone levels. With Dragon

Naturally Speaking, a user profile is created which includes the user’s age and accent type which will

affect speaking dictation. Dragon Naturally Speaking provides six different passages of text to register

speaking speed, accent and speaking tendencies to personalize the program to the user’s voice. This

process allows Dragon Naturally Speaking to have fewer mistakes than Windows Voice Recognition

software. With Dragon Naturally speaking, there is the ability to use dictation commands such as select

“word” and choose an option to correct the word without having to retype or re-speak the correction.

Page 7: Voice Recognition and Mobility in the Legal Industry

2

Also, Dragon Naturally Speaking allows the user to speak more freely and is able to translate words to

text faster than Windows Voice Recognition. The user would attempt to speak a few sentences using

Windows 7 Voice Recognition software resulting in less text correctly transcribed compared to using

Dragon Naturally Speaking. Dragon Naturally Speaking has more commands and better accuracy than

Windows 7 Voice Recognition, in part due to the Dragon Naturally Speaking User Profile that is built the

first time the program is used. The profile is constantly updated to improve accuracy through scheduled

recognition tasks, reviewing existing documents and a feature which allows for vocabulary

personalization.

Accuracy

Both programs had high accuracy percentages based on the 1,642 word narrative. Dragon Naturally

Speaking performed with 95% accuracy recognizing 1,575 words correctly, and Windows performed

with 93% accuracy recognizing 1,532 words correctly. Despite the relatively low 2% difference in

accuracy between the two programs, reviewing the Windows dictation was difficult as when it

incorrectly identified words it seemed to do so for strings of words making it difficult to understand the

complete sentence. Two examples of this from Appendix 1 are “Along the way they saw an Ice Cream

parlor” which both programs did not fully recognize as well as “Tom had been to many Phillies games

before.” Dragon mistakes were isolated from sentence to sentence which made it easier to understand

the narrative with the misspelled words. This is an important distinction since when reviewing a

document, the context of the sentence is important when trying to determine misspelled words. As

previously mentioned, the need for review cannot be overlooked, even if accuracy consistently performs

at 99%. In addition to the voice recognition results it was difficult recording with Windows Voice

Recognition as every sentence had to be read one by one and still some words were omitted. Also, the

Windows Software would make a loud “thud” if the text box was at capacity or if the program failed to

recognize a majority of the sentence being dictated.

Common Mistakes in Both Programs

The following section reviews the common mistakes both Dragon Naturally Speaking and Windows

Voice Recognition made. Some of the common mistakes both programs made were recognizing not only

certain words, but the context of these words. These examples show that while letting a computer do

the tasks of dictation and transcription which historically have been accomplished manually, the human

element of proofreading is still critical to document completion and must not be overlooked.

Both Dragon Naturally Speaking and Windows VR did not get the title correct.

Spoken: “A Day at The Ballpark”

Results:

Dragon Microsoft

I day at the ballpark I bit the ballpark

Page 8: Voice Recognition and Mobility in the Legal Industry

3

There are instances in both programs when what was dictated was not a complete sentence and

Microsoft Word did not flag any incomplete sentence structures with a subject, verb or predicate.

Spoken: “Along the way they saw an ice cream parlor”

Results:

Dragon Microsoft

Along the way this on ice cream parlor Along the way they saw ice cream parlor

When Dragon Naturally Speaking and Windows Voice Recognition do not recognize a word, it will

generate a similar word.

Spoken: “father and son day”

Windows Result: father and Sunday

Spoken: “Tom had been to other Phillies games before”

Results:

Dragon Microsoft

Tom had been to filter Phillies games

before

Tom had been too tough to lose games before

Test Results

Based on the testing and analysis of Dragon Naturally Speaking and Windows Voice Recognition, the

product used for the remainder of this paper will be Dragon Naturally Speaking. Based on testing Dragon

had the highest accuracy rate at 95%.

Before dictation and transcription models can be examined, the following section provides an overview

of the tools used to create and organize case-related documents. The creation of documents through

document templates will be reviewed using case-related data through the process of a document

merge.

Page 9: Voice Recognition and Mobility in the Legal Industry

4

Document Management Systems and Enterprise Content

Management

In the legal setting, Document Management Systems (DMS) and Enterprise Content Management

Systems (ECMS) are software that allow for the organization and easy retrieval of case-related data. The

term data is used broadly to cover legal documents, inter-office memorandums, internal and

external (sent and received) emails, voicemails and billing invoices.

These systems can scale from a solo practice consisting of a small workgroup of users to the top 100 law

firms in the United States with thousands of users, which gross more than one billion dollars, known as

AM100 law firms. As systems get larger, the configurations become more complicated. Searching for

data needs to remain a quick process as well as security concerns need to be addressed. While DMS and

ECM systems strive to solve the same problems and provide access to data, ECM systems take the

process further.

As the Association for Information and Image Management (AIIM) states, ECM focuses on the

strategies, methods and tools used to capture, manage, store, preserve, and deliver content and

documents related to organizational processes. ECM tools and strategies allow the management

of an organization's unstructured information, wherever that information exists (AIIM – What is

ECM?).

Document Management Systems will be the software of example used throughout the paper.

Document Management System Features

A Document Management System can consist of features such as a Document Repository, Document

Imaging, Multiple user access and Document Creation Templates (Davis, 2014). All case-related

documents are saved in a document repository within the DMS. This helps users access and collaborate

on documents as well as provides a secure location for client or case-related documents. A “specific

location” refers to a folder structure named by case number, case parties, or both. Since not all

documents received are electronic there is a volume of material received by the post office, overnight or

ground transport and courier as well. This material needs to be treated the same as electronic

documents and imaged by way of a scanner and copied into the DMS. The process for this is to scan all

documents and “import” into the DMS. Scanned documents can be converted to different formats

depending on the mechanism and software doing the formatting, but PDF is a universal format that is

used. Many of outgoing documents can have a designated format. Templates can be created which

quickens workflow and provides a uniform structure to all documents, regardless of content. For court

related documents, they need to be formatted in a style and/or layout each court dictates as acceptable.

For letters that go out to clients, opposing counsel, or any other parties related to a case, the outgoing

documents need to have a standard look throughout the entire document. The specific format makes

each document easy to read and creates brand awareness with the addition of a logo, office locations,

and firm web site. These documents are stored in a DMS which provides quick access and are accessible

by multiple users at the same time, regardless of geographic location. Some DMS systems will allow for a

check-out/check-in process so multiple users can view and edit documents at the same time creating

multiple versions once checked back in. In these systems the document is copied to the local machine

for edited. Once completed, they are copied back to the DMS as a new version. Other systems will mark

Page 10: Voice Recognition and Mobility in the Legal Industry

5

a document as read-only once subsequent copies are attempted to be opened. In this case a system can

provide options for additional viewers to be notified once the document is closed or to open as a new

version.

Document Management System Tools

Some of the DMS tools which are useful to either generating content or securing content are Security

features, Portable Document Format (PDF) functionality, Optical Character Recognition (OCR), Metadata

and full text searching and Document Scrubbing.

Security

As a DMS gets larger with various types of cases, security needs to be in place to limit document access

to only those who have the rights and privileges set by the Informational Technology Administrator.

Security groups can be setup based on the type of law the user works in. This is important since medical

records, personal financial information or intellectual property need to be protected from users that do

not need access to such content. Similarly related, Health Insurance Portability and Privacy Act (HIPPA)

and Freedom of Information Act are two relevant laws set by the United States Federal Government to

regulate the amount of information that can be disclosed about private citizens. There are nine

exemptions to disclosing information under the Freedom of Information Act. According to 5 U.S.C. §

552(b)(6), one of the nine exemptions is documents which are "personnel and medical and similar files

the disclosure of which would constitute a clearly unwarranted invasion of personal privacy”

(FOIAdvocates - Freedom of Information Requests/Appeals/Litigation).

Portable Document Format

Documents generated and sent out need to be protected from manipulation. A universally accepted

document type is the PDF. While standalone products like Adobe Acrobat Standard or Nuance PDF

Converter can convert documents to PDF, it’s helpful to have the conversion process built into the DMS

and it can be cheaper than licensing individual copies of software per users.

Optical Character Recognition

OCR is the process to convert a scanned image into readable text. This is important in a DMS to ensure

that when performing a full text search imported documents are searched as well as the internally

generated documents.

Metadata and Full Text Search

Metadata is useful for searching document properties instead of searching document content.

“Metadata is structured information that describes, explains, locates, or otherwise makes it easier to

retrieve, use, or manage an information resource. Metadata is often called data about data or

information about information” (Understanding Metadata, 2004). When creating documents,

information about the document can be used to find work by a certain author, or types and subtypes of

documents. A metadata search can be performed for documents that contain a certain number of

words, changes to text within documents, or even the type of operating system and word processing

Page 11: Voice Recognition and Mobility in the Legal Industry

6

software used to generate the document. Combining a full text search with a metadata search can

provide results based on content and format and/or author. A full text search is possible when the DMS

has a SQL process to index all documents for fast search, known as a Full Text Index. As the Full Text

Index runs, it indexes all character based columns in each table and stores them for future searching

(Full-Text Search, 2014). The index has to constantly run as documents are always being created and

modified. This is a function related to relational database software, such as Microsoft SQL or MySQL,

which a DMS is generally built on - it is not necessarily a function of the DMS itself, although the DMS is

the tool used to access the full text searches.

Types of Formatted Documents

Memorandums, status reports, and pleadings are just a few of the many types of documents lawyers

need to generate to meet client needs. As demonstrated later in the Appendixes, these documents can

completely be created and emailed via voice recognition. Status reports and pleadings have specific

formatting which a trained legal assistant can create faster than an untrained lawyer. The idea here is

with the proper merge document and the use of voice recognition software, the attorney can create

various status reports and pleadings without the use of a legal assistant. The benefits are reduction in

overhead including office space, fewer salaries to pay and faster work output – which is addressed later

in this paper.

Merge Document Example

As previously mentioned, the DMS is used to merge case data with a document template to quickly

create a standard formatted document as part of the work product.

The image in Figure 1 is a breakdown of an Interrogatory document template. An interrogatory is a

document which is sent to a party requesting information. The template consists of both static and

dynamic data. The dynamic data is referred to as “merged” data. By utilizing templates and merged data

the document generation process is streamlined. The need to search for, and enter data is eliminated as

is the need to manually format the document into its proper layout. Utilizing a template also eliminates

the need to repeatedly type standard language that is commonly generated and sent out. In the below

image, red highlight represents a dynamic merge code, green highlight represents the static template

and blue represents the static firm data.

Page 12: Voice Recognition and Mobility in the Legal Industry

7

Figure 1 – Interrogatory Template

The merge fields pull both case related data as well as firm data into the document from the DMS. Firm

static data is also merged. Instead of having multiple merged docs for various locations, only one

template needs to be created. When the data for a case is entered into the DMS, the data pertaining to

the case is used for the data merge. In the example, the Law Firm address in the upper left corner is

merged based on the office location of the attorney assigned to the case. Figure 2 shows the merge

codes highlighted in yellow. Merge codes consist of specific characters which make up a reference to a

Page 13: Voice Recognition and Mobility in the Legal Industry

8

field in the DMS database. In the example below, the first code (in red) is <<Assigned_Atty_full_name>>

which will pull the full name for the assigned attorney to the particular case. These codes are built into

the DMS by the provider and can be used as a reference in Word Processing programs such as Microsoft

Office or WordPerfect as well as certain version of Adobe Acrobat. Again, in the below image, red

highlight represents a dynamic merge code, green highlight represents the static template and blue

represents the static firm data.

Figure 2 - Merge Codes

Page 14: Voice Recognition and Mobility in the Legal Industry

9

Document Merge Steps

The document merge process is completed prior to transcribing dictation, or using Dragon to dictate

directly into a document. The six steps below in Figure 3 function the same regardless of the method of

dictation. Figure 3 is an overview of the steps a user will take upon determining which case information

needs to be merged. This is before the legal assistant starts typing or the attorney starts using Dragon

within the document. The six step process, as demonstrated in the flow chart and images below

illustrate a step-by-step process of using a DMS to create a new document. The overall process is

searching for a template and merging the template with data to create a formatted document ready for

content to be added.

In this specific example a legal assistant is using legal software from a company called Aderant. Total

Office (ATO) is the software used to create the documents. As previously stated, templates within a DMS

can be used to streamline document generation. This section explains the steps a user goes through to

launch and merge a document within the DMS.

Figure 3 - Document Merge Steps

Page 15: Voice Recognition and Mobility in the Legal Industry

Document Merge Process

Below is the step-by-step guide of merging a document from a template stored

letter. This is an example of the work

preparation of their case, such as contacting an insurance adjuster for a workman’s compensation case

or a civil court proceeding. In the example,

template description. This is a method used to find a template specifically built for sending a document

to an insurance adjuster. As each document is created, it is given a unique Document ID.

Step 1. User launches New Merge Document

Figure

10

guide of merging a document from a template stored in a DMS

an example of the work a legal assistant does on a daily basis to aid in the

contacting an insurance adjuster for a workman’s compensation case

civil court proceeding. In the example, the legal assistant is searching for adjuster wi

This is a method used to find a template specifically built for sending a document

As each document is created, it is given a unique Document ID.

Document from within a specific case in the DMS.

Figure 4 – Launch New Merge Document

in a DMS to draft a

on a daily basis to aid in the attorney’s

contacting an insurance adjuster for a workman’s compensation case

is searching for adjuster within the

This is a method used to find a template specifically built for sending a document

As each document is created, it is given a unique Document ID.

Page 16: Voice Recognition and Mobility in the Legal Industry

Step 2. Search for the template to be merged, launch selection.

Figure

11

be merged, launch selection.

Figure 5 - Template Search and Launch

Page 17: Voice Recognition and Mobility in the Legal Industry

Step 3. Select Party information to be used.

12

Select Party information to be used.

Figure 6 - Select Parties

Page 18: Voice Recognition and Mobility in the Legal Industry

Step 4. Make any necessary changes to recipient or from

ready select Merge.

Figure

13

Make any necessary changes to recipient or from lines, document type or add any notes. When

Figure 7 – Merge Document Summary and Review

, document type or add any notes. When

Page 19: Voice Recognition and Mobility in the Legal Industry

Step 5. Some merge documents will require additional

delivery method. E-Mail is the method

(and a blurred firm logo).

Figure 8 -

14

Some merge documents will require additional information; this letter needs to know the

Mail is the method which is manually typed. Notice the header consists of

- Additional Non-Merge Document Information

ds to know the

otice the header consists of codes

Page 20: Voice Recognition and Mobility in the Legal Industry

Step 6. Final merged document

15

Figure 9 - Final Document

Page 21: Voice Recognition and Mobility in the Legal Industry

16

Generating Content

Next is an examination of the current process a legal assistant goes through when transcribing dictation

using a digital dictation system and the DMS. After the examination of the current process, a

breakdown of two alternatives using voice recognition will be examined. These two alternatives are

using Dragon Naturally Speaking on a PC to generate documents and using a mobile solution with

Dragon software transcribing dictation on the server side and sending completed voice recognition files

to legal assistants for review.

Current Digital Dictation System Process

The current process, shown below in Figure 10, relies on an attorney to dictate to a handheld device,

which initially stores the dictation on its memory card. Once in the office the attorney can dock the

device to their computer which initiates the upload process and sends the dictation to the dictation

system queue. This is the Dictation Database shown on Figure 10. Prior to uploading, the attorney can

add notes to each dictation to specify which case the dictation is for or any special instructions. A level

of priority and/or a due date can be added as well.

The legal assistant accesses the attorney’s dictation queue from the transcription software which

launches at startup/login. Once the legal assistant reviews the dictation notes, the next step is to go

into the appropriate case within the DMS and launch the necessary merge template. Once the

document merge has completed, the playback of the dictation can start as the legal assistant transcribes

the recording by typing into the word processing software. The image below breaks down the process

which starts when an attorney uploads their dictation to the system, making it available for

transcription. The transcription is stored in a SQL server along with case related notes and any special

instructions. Once the dictation is available for transcription, the legal assistant will see it in their queue.

The information in the queue displays the case information and the type of document to be generated

so the legal assistant knows which case to open and which template to use in the DMS. After the legal

assistant transcribes the dictation the document is sent to the attorney for review. The attorney can

approve the document and add a signature to be sent out, or the attorney can request changes to the

document. If changes are necessary the legal assistant makes the changes and then submits the

document to the attorney for review. The cycle will continue until the document is ready to be signed

and sent out.

Page 22: Voice Recognition and Mobility in the Legal Industry

17

Figure 10 - Digital Dictation Process

Page 23: Voice Recognition and Mobility in the Legal Industry

18

Dragon on the Desktop

The second method used to add text to a document as generated in Figures 3 through 9, and the first

using voice recognition, is to use Dragon Naturally Speaking on the desktop. With this method, first an

attorney completes the document generation process on their own. Then, after launching the merge

document through the appropriate case within the DMS the attorney can use Dragon to dictate directly

into the document and bypass the traditional method of dictation/transcription with a legal assistant.

Using built in Dragon correction options, as well as the mouse and keyboard, the attorney can edit the

document as they dictate or review the document once initial dictation has been completed. Since there

is no recording of the dictation, it is not practical to send the dictated but not reviewed document to a

legal assistant to make corrections. It can be difficult for the legal assistant to make corrections for

words that are misspelled as well as words that are correctly spelled but incorrectly recognized by

Dragon.

Once the attorney has finished the dictation, there are two options. Depending on the content and type

of document, the attorney can email the completed document out. If it needs to be formatted a certain

way that wasn’t part of the merge template the attorney can have the legal assistant review. If there is

no need for review, and it must be sent via fax or regular mail, the legal assistant can do this. Since the

document is in the DMS, there is no need for the attorney to send the document to the legal assistant

via email and have them open, save, review, save, and attach back to the attorney. All the attorney

needs to do is provide the document ID and the legal assistant can update the document directly from

the DMS. Once done, the legal assistant can notify the attorney that the review has been complete, and

from there the document can go out. If a signature is required, it can be applied electronically through

various methods.

Figure 11 - Attorney Using Dragon on the Desktop

Page 24: Voice Recognition and Mobility in the Legal Industry

19

Dragon via Mobile

The third method used to add text to a document as generated through Figures 3 through 9 and second

using voice recognition is to combine the use of software to send dictations from a mobile device with

Dragon Naturally Speaking running on the backend as a software service. By using software from a

company called Big Hand, the attorney can dictate to, and send from the phone over either cellular or

Wi-Fi. The mobile software has the same functionality as the desktop software, providing relevant case

information the attorney can enter prior to sending.

As shown below in Figure 12 once the dictation is received by the Big Hand server, Dragon processes the

file for voice recognition and adds the completed file to the legal assistant’s queue.

Figure 12 - Attorney Using Dragon on the Server

When the legal assistant opens the voice recognition file, the transcribed text is viewed while listening

to the dictation. As each word is played back, Big Hand highlights the word on the document. This is

important for two reasons:

Page 25: Voice Recognition and Mobility in the Legal Industry

20

1. If the voice recognition fails to properly recognize the word and substitutes a similar word that

also makes sense within the context of the dictation, the legal assistant can catch the error.

2. If the voice recognition fails to properly recognize the word and instead adds a word that makes

no sense within the context of the dictation, the legal assistant can catch the error.

Once the voice recognition has been reviewed, and the document has been properly formatted the

document options are the same as previously described; the document can be sent out as is, or it can be

reviewed by the attorney prior to being sent out.

Below is a screenshot of the view the legal assistant has when playing back dictation that has been

recognized by Dragon through Big Hand. Notice the word “advised” highlighted in blue as it was the

active word being played back at the time of the screen shot. This is step “legal assistant reviews text,

makes changes using Big Hand MS Word add-in” previously shown in Figure 12.

Figure 13 - Screenshot of Dictation Playback with Voice Recognition

By utilizing both Big Hand mobile software and Dragon on the server side, attorneys can dictate and

send from anywhere as long as they have a cellular signal or are connected to Wi-Fi. By using the mobile

application there is no need to purchase new dictation devices such as USB microphones or mobile

recording devices which can cost up to $450 each. Using the mobile application also eliminates the need

to purchase Dragon software (at a minimum $100) for the PC or Mac Desktop for those attorneys that

don’t want to use the software to enter time, draft emails, or for any other functions other than

dictation. Also, when using Dragon on the server side, voice profiles are stored on a networked server

and backed up. Local Dragon installs do not have this functionality and the profile is lost if the computer

is subject to a catastrophic event. And one final point, voice recognition processing is done on the server

Page 26: Voice Recognition and Mobility in the Legal Industry

21

side which does not impact performance on the user computer or other applications as running Dragon

on the local computer can do.

Analysis of Dictation Workflow

Dragon on the Desktop

Three methods of dictation have been reviewed.

1. Traditional dictation and transcription

2. Dragon on the desktop

3. Dragon on the server side

By focusing solely on the workflow the quickest way to have completed dictation is for the attorney to

use Dragon on the desktop. The completed process with Dragon on the desktop consists of the least

amount of steps. Dragon streamlines the document generation process and eliminates the need for

another user to review and cleanup the document which can reduce overhead and facilitate a faster

process to get documents out the door quicker. Eliminating the document approval process shown

below in Figure 14 can drastically save time since every change needs to be reviewed which can cut into

the attorney’s work day. The document approval process occurs in both the Digital Dictation Process

(Figure 10) and Attorney Using Dragon on the Server Process (Figure 12).

Figure 14 - Document Approval

Attorneys dictating using Dragon directly to the computer have the ability to create and correct more

documents than the traditional attorney/legal assistant model. Dragon can process 160 words per

minute (WPM) once a profile has been correctly built over time. It’s challenging to find documentation

for a legal WPM standard, but the average of 70-80 WPM is backed up (Arif, 2009) and acceptable for

the use of comparison.

If an attorney is dictating on average 160 WPM to Dragon, on average they are speaking a total of 9600

words per hour. A legal assistant that averages 75 WPM will type an average of 4500 words per hour.

Page 27: Voice Recognition and Mobility in the Legal Industry

22

Dragon corrections can take between four and 15 seconds per correction (Alexandria, 2011). If using the

average of 10 seconds per correction (the split between 4 and 15 seconds rounded up) as well as the

example of 67 corrections needed to be made after dictating a 1,642 word document the result is 388.6

corrections for a 9600 word document. The voice recognition testing took roughly ten minutes, so the

number of corrections in the example is accurate. 388.6 corrections will take an average of 38.9

minutes. By adding the additional time to an attorney’s 8-hour day, they dictate for an hour and correct

for 40 minutes with a result of 48,000 completed/fixed words per day. A legal assistant typing 8 hours

per day at 75 WPM corresponds to 36,000 words before any proofreading is done.

Put another way:

1,642 words dictated at 160 WPM (10.2 minutes) + correcting 67 errors at 10 seconds each (7.10

minutes) = 17.3 minutes. The charts below display same.

Words Dictated Words Per Minute

(Dictated)

Time to Dictate (Minutes)

Attorney 1,642 160 10.2

Number of Errors Time to Fix Each

Error (Seconds)

Correction Time

(Minutes)

Final Time

Including

Typing

Corrections

(Minutes)

Attorney 67 .10 7.10 17.3

1,642 words typed at 75 WPM = 21.8 minutes

Words Dictated Words Per Minute

(Transcribed)

Final Time to Transcribe

(Minutes)

Legal Assistant 1,642 75 21.8

The example of 67 errors is from initial use. As accuracy improves, those 67 errors (per 1,642 words) will

likely decrease over time. Dragon will never be 100% accurate. If accuracy can improve and reduce the

errors in half to 34, the time to complete a document using the example will decrease from 17.3

minutes to 13.6 minutes. Chart below displays same.

Words Dictated Words Per Minute

(Dictated)

Time to Dictate (Minutes)

Attorney 1,642 160 10.2

Number of Errors Time to Fix Each

Error (Seconds)

Correction Time

(Minutes)

Final Time

Including

Typing

Corrections

(Minutes)

Attorney 34 .10 3.40 13.6

Page 28: Voice Recognition and Mobility in the Legal Industry

23

Measuring the Success of Dragon on the Desktop

Using Dragon on the desktop is an individual software solution and not an enterprise software solution.

Enterprise software solutions can provide usage reports and tools to manage and monitor the software

from a higher level. The use of Dragon on the desktop has no “higher level” capabilities. One of the

drawbacks of this is it is difficult to measure or evaluate its success. Success for Dragon on the desktop is

best evaluated by firm management from the top down. Partners that have a stake in the business side

of the firm need to use Dragon on their own as a means to lead by example. These partners also need to

enforce the use of Dragon on the desktop through the mentoring of associates. Associates need to

understand the financial benefits of Dragon and partners need to use monetary incentives to enforce its

use. It needs to be made clear to the associate that by relying less on the legal assistant and building the

business without additional staff can lead to an increase in salary for the associates as well as partners.

Lastly, an attorney that is evaluated yearly with an annual review can be reviewed on their contribution

to less overhead by using Dragon, in addition to other attorney review staples such as billable time and

client development.

Measuring Return on Investment of Dragon on the Desktop

Using Dragon on the desktop ultimately allows for a higher ratio of attorneys to legal assistant. Assume

the baseline ratio is two attorneys to one legal assistant. With attorneys doing most, if not all of their

dictation, it is possible to increase the attorney to legal assistant ratio to four to one. For every two

attorneys a legal assistant can support, it is one less legal assistant a firm needs to hire. This means a

firm can grow the business and hire additional attorneys without needing to hire additional staff,

creating less need to spend at a minimum $56,750 for a new legal assistant.

Salary $55,000

Benefits Varies

Office space Varies

Computer equipment $1,000

Software Licensing $500

Bandwidth Varies

Phones and phone lines/extensions $250

Total $56,750

Dragon via Mobile with Voice Recognition on the Server Side

As for the other two methods discussed, traditional dictation and transcription and Dragon on the server

side via mobile, the mobile process using Big Hand software can increase dictation efficiency as well. In

those situations when an attorney is travelling and without the ability to send dictation to their legal

assistant, the mobile app will prevent an attorney from having to go out of their way to get back to the

office to upload the dictation. The mobile app also allows the attorney to continue to dictate without

creating a backlog of work and helps facilitate a steady stream of work for the legal assistant who might

find themselves with less work when the attorney is out of the office for extended periods of time. Of

the three methods, Dragon on the server side via mobile is a highly effective process, despite the extra

Page 29: Voice Recognition and Mobility in the Legal Industry

24

steps needed for review as shown in Figure 12. The mobile option ensures that an attorney’s work will

continue to be transcribed and/or reviewed regardless of where the attorney is working from and it’s

the best option when Dragon on the desktop isn’t available.

With using voice recognition on the server side the process shifts for the legal assistant from a creator

role to a reviewer role. In this reviewer role the Legal assistant needs to focus on mistakes that are

already made by voice recognition, which is a shift in how they are traditionally trained but not too far

removed from their overall responsibilities. The shift from traditional transcription to document

reviewer does not require any extensive training nor are there any hidden costs associated. The training

to use the Big Hand add-in is minimal as it is an easy process to click play and listen back to the

document while reviewing. As for hidden costs, Big Hand only charges a Dragon license for the dictation

component, not the transcription side. For example, a firm with five attorneys and 10 legal assistants to

handle the review would only require five Dragon licenses.

After a legal assistant transcribes from dictation the document should be given a “once over” to

proofread prior to sending back to the attorney for review. This is a “best practice” approach that

applies to everyone in any industry or setting that relies on document generation and exchange.

Once the finished text is reviewed, the process to complete the document begins with a new merge

process and ends with sending the document back to the attorney for review. Below in Figure 15, the

voice recognition review process is reiterated. While using Big Hand to process voice recognition is an

efficient process, one of the drawbacks is the reliance on typists to become reviewers either in addition

to or instead of typists. While some Legal assistants can “go with the flow” and adapt, others that are

more resistant to change could provide some “pushback” as a defense mechanism to the fact that they

aren’t typing anymore and that a software application could replace them. This is an unfortunate

criticism of the individual and not a reflection of the software. Obviously, with the outlined process it is

not entirely true but the reality is that with attorneys using Dragon on the server side to transcribe their

dictation, the text review process takes less time than traditional transcription so more attorneys can be

assigned to a legal assistant. This knowledge can instill fear in the legal assistant which can result in

“pushback”. The term pushback is used here to describe a legal assistant not adapting to change and

speaks to an overall theory of the use of the term “legal assistant” compared to “secretary”. When using

the term legal assistant, it can be assumed that the support being provided is more than just a standard

typist/transcriptionist. The legal assistant needs to perform other duties such as court filings and

maintaining schedules. There are legal assistants that only act as secretaries – they just want to type

dictation and everything else related to the job is a chore. If you remove the primary function of

transcription and make it a job to review then the good legal assistants understand the value, take on

more attorneys and can work as a true assistant. The ones that pushback and resist change complain

about the new process and don’t adapt. Once pushback begins the legal assistant could be moved to a

pure secretary position if available. In some cases the legal assistant could be considered for termination

and in response the attorney whom the legal assistant works for can step in and tell firm management

that despite the issue, due to a positive working relationship no changes should be made and the

attorney can resist the use of voice recognition.

Page 30: Voice Recognition and Mobility in the Legal Industry

25

Figure 15 - Dragon Review

Big Hand Costs

Big Hand is an enterprise solution, and like most enterprise solutions pricing is per user and per

component, along with a server licensing expense as well. Pricing includes support, and since Big Hand is

server/client based, support is critical.

A Big Hand user license is $495 and the Server license is $5,495. Every user, regardless of author

(Attorney) or assistant (Legal assistant) requires a license. Hourly on-site support is $225 sold in blocks

of 8 hour days. Options for on-site support, which is all billable, include project management, technical

implementation and training. Once the implementation is completed, support is handled by remote

access, via email or by phone. Annual maintenance, which includes support, is 18% of the total system

price.

Using Dragon on the server side via the mobile application requires additional licensing. The mobile app

is free, but each user license is $150. The mobile app provides the ability to send traditional dictation

direct to the attorneys queue or to the Dragon server which processes the voice recognition before

adding to the attorney queue. While the standard app will do either, a license for Dragon on the server

side is required per attorney at $595 each. Additional licensing costs are Microsoft SQL server and

Microsoft Windows server licenses. Voice recognition requires its own server to process the files, so at a

minimum two servers (hardware) as well as server licenses are required. These costs can vary due to a

law firms existing licensing agreement with Microsoft as well as physical vs virtual server

implementations.

Page 31: Voice Recognition and Mobility in the Legal Industry

26

Big Hand Vs Dragon Cost Breakdown

Big Hand User License $595

Big Hand Server License $5,495

Hourly Support $225

Big Hand Mobile User License $150

Dragon Server Side User License $595

Annual Maintenance 18%

Dragon Naturally Speaking $199

Using the example of a small firm with 25 attorneys and 10 legal assistants, the initial cost to implement

Big Hand is $57,283 before any server expenses.

Each Licenses Total

Big Hand User License $595 35 $20,825

Big Hand Server License $5,495 1 $5,495

Hourly On-Site Setup Support $225 16 $3,600

Big Hand Mobile User License $150 25 $3,750

Dragon Server Side User License $595 25 $14,875

Subtotal $48,545

Annual Maintenance 18% $8,738.10

Total Year One $57,283

The cost to implement Dragon Naturally Speaking using the same size firm is $4,975.

Each Licenses Total

Dragon Naturally Speaking $199 25 $4,975

Compare the costs associated with Big Hand with the cost to purchase Dragon on the desktop, and it’s

easy to justify Dragon on the desktop, especially for a solo or small practice.

Dragon Naturally Speaking Implementation

Hardware Requirements and Cost

If a user is serious about using voice recognition to be more efficient, then an investment in the basic

version of Dragon Naturally Speaking Professional should be made. You get what you pay for is the old

adage, and that applies to the free version of Windows Voice Recognition.

Dragon Naturally Speaking Premium 12 retails directly from Nuance for $199.

Page 32: Voice Recognition and Mobility in the Legal Industry

27

There are numerous Microsoft operating systems which Dragon Premium 12 can run on. They include

Microsoft Windows 7 and higher, 32-bit and 64-bit; Microsoft Windows Vista SP2, 32-bit and 64-bit;

Microsoft Windows XP SP3, 32-bit only; Windows Server 2008 SP2 and R2, 32-bit and 64-bit.

Dragon recommends having 2GB RAM for Windows XP, Windows Vista, Windows 7, and Windows

Server 2008 32-bit and 4GB for Windows 7 and Windows Server 2008 64-bit. The minimum system

requirements are 1GB for Windows XP and Windows Vista, and 2GB for Windows 7 and Windows Server

2008.

Dragon also recommends a 2.2GHz Intel dual core or equivalent AMD processor. The minimum

requirement is a 1GHz Intel Pentium or equivalent AMD processor or 1.66 GHz Intel Atom processor. For

the processor cache, 2MB L2 cache is recommended, with the minimum requirement at 512KB L2

cache. In addition to RAM and CPU specifications 3.2GB of free hard disk space (or 4GB for localized

non-English versions) is necessary (Duffy, 2012).

Our tests were conducted using Windows 7 64-bit with 4GB of RAM and an Intel Dual Core i5-2410M

CPU @ 2.30GHz.

Installation and Training

Given the ease with which Dragon streamlines the installation process there is no need for small law

firms without a dedicated IT staff to outsource the installation. Since each user needs to be involved

with setting up the voice profile, they can be part of the install process as well. The guided install and

profile setup can be completed by anyone regardless of technical ability.

An attorney at a larger firm attempting to install Dragon on their own could have an issue due to a firm

policy that has locked down the computers on the network. In this situation coordinated install would

need to be performed between the IT department and the user to first install the software, second

ensure the microphone is working and third, have the user complete the voice profile setup.

Once the setup has been completed Dragon goes through an overview which explains the various

commands to use while dictating; moving throughout a document, applying text styles and various

correction options. Also, shown below in Figure 16, Dragon makes use of a sidebar which changes as the

user changes applications. The sidebar will automatically show commands for Microsoft Word, Internet

Explorer and the Windows desktop as each application is made active. This sidebar is an excellent

resource as a user guide and reference as it automatically opens each time Dragon is launched.

Page 33: Voice Recognition and Mobility in the Legal Industry

28

Figure 16 - Dragon Sidebar

Page 34: Voice Recognition and Mobility in the Legal Industry

29

Voice Recognition vs Manual Transcription Pathology Lab

Study

According to the Medical Laboratory Observer from September 2012, a surgical pathology lab

performed a study to determine if replacing manual transcription with voice recognition software was

cost effective and created a series of tests to determine the effectiveness of voice recognition compared

to manual transcription. The study also focused on the annual costs of the voice recognition solution

and the Return on Investment (ROI) based on a payback period of at least five years.

The study used two comparable surgical pathology departments. One department used voice

recognition and the other used manual transcription with dictation.

Workflows for both voice recognition (VR) and manual transcription (MT) methods were created and a

Cycle Time (CT) for each component of the workflow was collected for all personnel. The timed studies

accounted for

• Pathologist Assistant Grossing

• Pathologists Microscopic Dictation

Once the CT’s were collected, a result of average cost per word (CPW) was created and equations were

used to compare both methods of VR and MT. By using salary.com to estimate average salaries of the

PA, labor costs were calculated per process using the below equations. The term “specimen” relates to

each completed histopathological report and is the basis for all analysis.

• Average Word/Specimen (total words for all specimen/number of specimens)

• Average Time/Specimen (total time for all specimens/number of specimens)

• Average Cost/Specimen (salary (per min) for each person performing a task x)

• Average Cost/Word (average cost/specimen / average word/specimen)

• Labor Cost for Process (average cost/specimen x number of specimens)

Based on these formulas it was determined that while completed times for finalized reports was less for

VR (2.21 min) than for MT (3.09 min), the overall cost for VR per specimen ($3.89) was significantly

higher than MT ($2.43). This resulted in a difference of $1.46 per specimen when using VR. The article

states that “the criteria for rejection is a calculated payback period of greater than five years, and in our

situation payback would never occur.”

This study does not address the added expense of hardware, software or training. The potential savings

related to hiring less staff, which can consist of salaries, benefits and office overhead is not factored as

part of the ROI. The ROI is based solely on the cost of time difference (per specimen) and that is not the

complete picture. As reports were completed on average 47 seconds faster using voice recognition, this

study confirms that voice recognition as a process compared to manual transcription can produce the

same efficiencies tested in this paper, and in a different industry. The study demonstrates that voice

recognition software solutions can span multiple industries, especially those that rely on traditional

dictation and transcription methods.

Page 35: Voice Recognition and Mobility in the Legal Industry

30

Conclusion

The classic combination of dictation and transcription has been in place for years because it was the only

method that worked. As technology increased, attorneys moved from dictating to tapes to dictating to a

digital source. Hardware for legal assistants continued to evolve as dictation moved from analog to

digital so transcription could continue as the processes changed. Dragon software has been around

since 1997, but not until the last few years has computer hardware caught up with what Dragon set out

to initially accomplish. It’s always worked, just not at an acceptable level to change the process – until

now. Because of this, Dragon is the future of document transcription and voice recognition.

Of the three proposed methods of dictation; traditional dictation and transcription, Dragon on the

desktop and Dragon on the server side via mobile, the selection of Dragon on the desktop as the most

efficient of the three could be viewed as somewhat controversial within the legal industry, especially for

those older attorneys that rely heavily on dictation. Shifting the responsibility from the legal assistant to

the attorney for document generation is controversial due to the change in process, responsibility and

thinking. The change in process is an attorney using Dragon would need to sit at their computer and

dictate instead of standing or walking around while dictating. If dictating includes referencing

documents, the attorney would need to get used to the task of referencing documents either printed or

electronic and then switch to the active document that is being dictated to. The most controversial

aspect could be for those older attorneys that have struggled to accept how technology plays an integral

role in the practice of law. Attorneys that refuse to keep an electronic calendar, that don’t know how to

type, or those that refuse to learn the basics of firm software, such as a DMS, are the ones most likely to

not accept the transition from traditional dictation to using Dragon on the desktop. In these cases, it

must be understood that if an attorney is refusing to use Dragon and not taking the necessary steps to

reduce overhead in the form of hiring less legal assistants, then the attorney could be penalized

financially in the form of less bonus money, or less profit sharing if the attorney is a firm shareholder.

When examining the workflow of all three solutions Dragon on the desktop is the most time efficient

way for an attorney to generate any type of document that is traditionally dictated and transcribed. It is

the only workflow that does not rely on the repeated process of document review between attorney

and transcriptionist. It is the only workflow that provides a completed document to the legal assistant to

be sent out. As mentioned earlier, with a properly configured Dragon voice profile the software can

recognize 160 words per minute which is more than twice as fast as a legal assistant typing 75 words per

minute. As Dragon is continually used it becomes more accurate. Dragon has built in voice recognition

commands and when learned and used will contribute to an attorney’s efficiency as they can recite

commands at the same time as speaking. This can greatly reduce the time to clean up and review once

dictation has been completed.

Dragon on the desktop is also the most cost effective method for an attorney to dictate a letter.

Attorneys can bill their time to draft the letter, as well as the time it takes to make corrections. Dragon

on the desktop reduces dependencies on staff and can cut overhead expenses as there is less need to

hire additional staff as firm business increases with the hiring of additional lawyers. Due to these two

examples, it is difficult to debate the $199 cost of the software.

Moving forward, Dragon Professional 13 is slated for release (August, 2014) and claims to be 15% more

accurate than version 12 (which is the version we used). Accuracy is the area Dragon will need to

continue to improve on and will end up being the game changer that law firm management needs to

Page 36: Voice Recognition and Mobility in the Legal Industry

31

take to the attorneys for wide spread implementation. As it becomes even more accurate it will

continue to gain market share, becoming the standard in voice recognition on the desktop. By doing so

Dragon could potentially eliminate the old dictation/transcription model not just in the legal profession,

but in any other professions that follow the model. Future review of this project will be to continually

evaluate new versions and upgrades of Dragon, along with advances in hardware, such as faster

processors. Each new release should be tested using the same standards as this paper. Attorneys should

collaborate on exploring new features, especially those smaller firms that do not have dedicated IT staff.

Lessons Learned

Initially, the idea of Dragon on the server side delivering completed transcription to a legal assistant

seemed like the obvious choice as the most efficient process of the three examined. The process

eliminated any typing, and provided a legal assistant the tools to listen back to the dictation and make

the changes that the server didn’t process correctly. As further examination occurred it became

apparent that despite the completed transcription the review process was going to take just as long as

the traditional method of dictation and transcription. Once that was realized, Dragon on the desktop

became the clear front runner as the most efficient process since the attorney could complete a

document on their own and eliminate the lengthy back and forth process of review.

Another lesson learned was how important correct voice recognition needs to be viewed within the

context of a sentence. One of the major reasons for choosing Dragon Naturally Speaking over Windows

Voice Recognition was due to not only did Windows incorrectly transcribe more words incorrectly, but it

would incorrectly transcribe more words in a row within a sentence making it much more difficult to

determine what was actually said.

As for the Return on Investment for Dragon on the server side via Big Hand, it has a place in the legal

environment, but maybe only in larger firms that can afford the expensive software as well as afford to

NOT enforce Dragon on the desktop. A larger firm that has a high number of older partners is most likely

going to resist the change from traditional dictation to using Dragon on the desktop. For these firms, the

move to Dragon on the server side via Big Hand could make sense. The larger firms can afford it, and

some of the “forward thinking” older partners can see the value in the software and use it to limit hiring

additional legal assistants when using the server side transcription component.

Page 37: Voice Recognition and Mobility in the Legal Industry

32

References

AIIM - What is ECM? What is Enterprise Content Management?. (n.d.). AIIM - What is ECM? What is

Enterprise Content Management?. Retrieved May 30, 2014, from http://www.aiim.org/what-is-ecm-

enterprise-content-management

BigHand Acquires Leading Legal Software Company Esquire Innovations, Inc.. (2014, March 19). .

Retrieved May 29, 2014, from

http://www.reuters.com/article/2014/03/19/idUSnMKW1SC5Za+1da+MKW20140319

Carstens, A. (2011, June 1). . . Retrieved July 11, 2014, from

http://www.speakeasysolutions.com/blog/2011/06/01/the-myths-realities-of-roi-in-implementing-

speech-voice-recognition-technology-dragon-in-law-firms/

CICCATELLI, A. (2014, August 15). BigHand speech recognition instantly produces client witness

statements. Retrieved August 27, 2014, from http://www.insidecounsel.com/2014/08/15/bighand-

speech-recognition-instantly-produces-clie

Davis, B. (2014, May 13). Document Management Software Review 2014 | Best Document Management

System - TopTenREVIEWS. TopTenREVIEWS. Retrieved June 6, 2014, from http://document-

management-software-review.toptenreviews.com/

Document Management Overview. (2007, January 1). Retrieved June 9, 2014, from

http://www.lhric.org/files/filesystem/Document Management Guide.pdf

Duffy, J. (2012, September 14). Dragon NaturallySpeaking 12 Premium. Retrieved July 15, 2014, from

http://www.pcmag.com/article2/0,2817,2409720,00.asp

Eaton, K. (2013, April 17). Make a Note of It: Speech Recognition Apps Are Getting Better. The New York

Times. Retrieved June 5, 2014, from

http://www.nytimes.com/2013/04/18/technology/personaltech/dragon-dictation-and-other-speech-

recognition-apps-review.html?_r=0

ECMWise: The Content Management Blog: Solving the puzzle: DMS, RMS or ECM. (2012, September 24).

ECMWise: The Content Management Blog: Solving the puzzle: DMS, RMS or ECM. Retrieved June 6,

2014, from http://ecmwise.blogspot.com/2012/09/solving-puzzle-dms-rms-or-ecm.html

FOIAdvocates - Freedom of Information Requests/Appeals/Litigation. (n.d.). Retrieved September 3,

2014, from http://www.foiadvocates.com/exemptions.html

Page 38: Voice Recognition and Mobility in the Legal Industry

33

Full-Text Search (SQL Server). (2014, June 20). Full-Text Search (SQL Server). Retrieved July 1, 2014, from

http://msdn.microsoft.com/en-us/library/ms142571.aspx

Houlihan, D. (2014, March 26). Will the Real Winner of BigHand's Esquire Innovations Acquisition Prove

to be Small Law?. Blue Hill Research. Retrieved May 29, 2014, from http://bluehillresearch.com/will-

small-firms-be-the-real-winners-in-bighands-esquire-innovations-acquisition/

Leikums, T. (2013, July 1). Document Management System Mobility – Main Trend in Future Document

Management. Retrieved July 17, 2014, from http://www.ijcce.org/papers/236-W0025.pdf

Magic Quadrant For Enterprise Content Management. (2012, January 1). Retrieved June 17, 2014, from

http://www.project-consult.de/files/Gartner Magic Quadrant for Enterprise Content Management

2013.pdf

North Dakota Information Technology Department (ITD). (n.d.). Retrieved June 14, 2014, from

http://www.nd.gov/itd/services/enterprise-document-management-system-edms

Suvalsky, Steven,M.H.S., P.A.(A.S.C.P.), & Boelman, Valerie,B.S., L.B.C. (2012). Voice recognition

technology vs. manual transcription. Medical Laboratory Observer, 44(9), 36-39. Retrieved from

http://search.proquest.com/docview/1356649001?accountid=11999

Understanding How Law Offices Do Business. (n.d.). Retrieved September 1, 2014, from

http://research.lawyers.com/understanding-how-law-offices-do-business.html

Understanding Metadata. (2004, January 1). . Retrieved June 6, 2014, from

http://www.niso.org/publications/press/UnderstandingMetadata.pdf

Voice technology up sharply by 2015. (2011). Material Handling & Logistics, , 9. Retrieved from

http://search.proquest.com/docview/868680799?accountid=11999

Page 39: Voice Recognition and Mobility in the Legal Industry

34

Appendix 1: Narrative

In this exercise, our team composed a narrative story about a common day when a father takes his two

sons to a baseball game in a word document using both Dragon Naturally Speaking and Windows Voice

Recognition. The first step was composing the script in Dragon Word Pad by dictating the story through

a microphone. Afterwards, the content was edited using Dragon Naturally Speaking and reread using

Dragon Naturally Speaking in Microsoft Word as well as Windows Voice Recognition in Microsoft Word.

Below is our original script with the unedited versions of the content from both programs, highlighted in

bold are the initial miss-spellings Dragon and Windows Voice Recognition made. Not all spelling and

grammar mistakes are highlighted, just a selection in both.

Dragon Naturally Speaking (Unedited)

I day at the ballpark

it was not a typical Sunday afternoon in the Wilson household. Mr. Wilson wanted to treat his two sons

Charlie and Tom to a beautiful day at the ballpark after finishing another school year. Charlie, and Tom,

age 12 always wanted to go to citizens Bank ballpark in Philadelphia PA, but have never gotten the

chance to go. Charlie, Junior at Central Bucks East high school, never got the opportunity to go to a

Philadelphia Phillies baseball game because he was so studious. Charlie’s younger brother Tom had been

to filter Phillies games before, but this time he was very excited to have a father and son day at his neck

ballpark. Mr. Wilson thought it was a great idea to take the septa train instead of driving to the ballpark

because it’s very hard to park at the stadium and parking is cheaper at the train station. Mr. Wilson,

Charlie and Tom got in the car and drove to the train station. Along the way this on ice cream parlor. Mr.

Wilson wanted to treat his two sons to an ice cream cone before the game because it was a very hot and

humid day. Mr. Wilson’s finishing flavor is mint chocolate chip, Charlie’s favorite is by recon, and Tom’s

favorite is cookies and cream. After ordering their ice cream Mr. Wilson and his two sons sat in nearby

picnic table and enjoy their ice cream cone while taking about who would be the starting pitchers for

today’s game. Charlie and Tom were very excited to spend time with their father because during the

school year they really do not have a chance to bond. After finishing their ice cream cone, the Wilson’s

got back in the car we continue driving to the train station. Once they got there, they boarded the

orange line, which is known as the Broad Street line. Charlie wanted to sit in the back of the train, but

his brother and father wanted to sit in the front. Solve this dilemma, Mr. Wilson flipped a coin. Charlie

chose heads, but the coin landed tails. Tom had won the coin toss, leading them to sit in the front of the

train. On the train there were a lot of Philadelphia Phillies fans, Beverly was very happy and anxious to

get to the ballpark early. The reason why everybody on the train wanted to get to the ballpark early was

because it was picture today with the Philadelphia Phillies and their mascot, the Philly Phanatic. Charlie

and Tom had taken a photography class in school this past semester wanted to show off their talents.

On the way the train had to stop at several stops as it made its way to a T and T station, which is the

transition for all Philadelphia stadiums. Once every got off the train, the road to the escalator up to

street level. As a went through the revolving doors, to their left they saw the beautiful and magnificent

citizens Bank ballpark. It was only a short walk to the ballpark and only along the way they saw street

musicians and street vendors. After eating their ice cream cone, Mr. Wilson, Charlie, Tom were very

thirsty. They decided to each by a bottled water to consume as they waited to get inside the stadium

area once they finish their bottled water it was time to open the gates and go inside. Mr. Wilson pulled

out the three tickets that he bought from stub hub the night before and gave one tickets each of his

Page 40: Voice Recognition and Mobility in the Legal Industry

35

sons. Before they enter the gate, Charlie wanted to buy a program with a scorecard to keep track of the

game. This is something that he always enjoyed doing when watching a baseball game at home. The

program was scorecard cost five dollars and Charlie pulled out a $20 bill that he earned for getting

straight A’s on his last report card. Her Charlie it felt so rewarding to work hard at school would be able

to spend his money this time with people he enjoyed being around. As Mr. Wilson, Charlie and Tom

were walking into the stadium after hitting their tickets to the stadium attendee, they saw the Philly

Phanatic. Charlie and Tom were very excited and waited in line to take the phanatics picture as a got to

the front of the line, one of the stadium attendee’s offer to take Mr. Wilson, Charlie, Tom’s picture

together with the Phanatic. As they continued down the hallway, they saw some of the Philadelphia

Phillies at a table signing autographs. There were former or current Phillies there, such as Mike Schmidt,

Jimmy Rollins, Greg Lubinski, and Dominic Brown, as well as the filter Phillies TV broadcasters Tom

McCarthy, Matt stairs, and Jamie Moyer. This was a thrill for any Philadelphia Phillies fan is everyone

waited patiently in line to get their picture taken and to get an autograph. Mr. Wilson, Charlie and Tom

were able to get everyone’s autograph their and had a permanent souvenir to treasure. As they were

finishing up in line they still had about an hour 50 minutes ago before the first pitch. Mr. Wilson, Charlie

and Tom were still hungry and wanted a ballpark Franks hot dog. The Wilson’s weaved in and out of the

crowd to get to the hot dog stand located at the above concourse. Once they got to the hot dog stand,

everyone had to choose what toppings they wanted. Mr. Wilson chose mustard, relish and onions on his

hot dog while his two sons chose catch up and onions. Mr. Wilson paid for the hot dogs which included

a soft drink and they made their way up to their seats one of the most fun things to do in going to a

baseball game is watching batting practice the visiting team, the New York Mets, were on the field

taking batting practice and fielding flyballs. With the binoculars that Mr. Wilson brought, he was able to

see all the players up close, even though their seats were 30 rows back from first base. For about

another 20 minutes, the New York Mets continue to take batting practice while the Phillies were whiting

to take the field. After the Mets went back to their dugout, the Phillies took the field. Mr. will supersede

Dominic Brown, Carlos Ruiz, Jimmy Rollins and Chase Utley waiting behind the batting cage. It was a

sunny day, about 80°, and was a beautiful day for a ballgame. For about 30 minutes, the Phillies

practiced her fielding and batting swings and it was time to change into their game uniforms. While both

teams are making last preparations before the game, on the Jumbotron they were showing highlights

for the previous week in Maj. league baseball. Mr. Wilson, Charlie and Tom were very fascinated with

the highlights and clapped after the video was over. The Philly Phanatic then came onto the field on his

four wheeler and drove around the entire infield waving to the fans. The Phanatic then went off on his

four wheeler and walked to the pitcher’s mound were two children were waiting to throw out the

ceremonial first pitch. Behind home plate were Philadelphia Phillies catchers Carlos Ruiz and will be

Ava’s. The two children through the first pitch with varying strength as a two catchers attempted to

catch the balls. After the two children throughout the pitch, the two Phillies catchers went to the

pitcher’s mound to greet the children and the Phanatic. The public address announcer from citizens

Bank ballpark announce that the Philadelphia squire would be singing the national anthem. All the fans

of the ballpark rose to their feet as a audit America afterwords, the two teams managers met at home

plate, truck hands and exchange batting lineups. The briefly discuss the rules of the ballpark with the

umpire at home plate. The game was about to begin the starting pitchers were Cliff Lee for the Phillies

and Jonathan needs for the Mets the first batter for the Mets was Curtis Granderson. After a four pitch

at-bat league ended up striking Granderson out. The following two hitters provided to be no match for

Lee, as a lot of the first quickly approached. Nice also brought his a game, and so the game quickly

became a pitcher’s duel with both Lee and Nice only giving up a few hits, both teams is offenses found it

difficult to score. It wasn’t until the bottom of the seventh inning when the Phillies finally put one up on

the scoreboard, thanks to Carlos releases sac fly that brought home Marlon Byrd. The Phillies hung on to

their one run lead long enough for their closer Jonathan Topol Kuan to calm in the top of the ninth and

Page 41: Voice Recognition and Mobility in the Legal Industry

36

close it out with two strikeouts and one groundout. The Phillies get a once enough victory at home Mr.

Wilson his two sons clap and cheer as a wanted to congratulate each other on their win. They excitedly

watch on the Jumbotron a recording of the late Harry Kalas singing high hopes as a ballpark crucially

plays after home victory. All this excitement and cheering seemed to stir up an appetite Mr. Wilson and

his two sons, so they decide to separate the victory the satisfying meal from Philadelphia’s very own

chickie and Pete’s. They all enjoy their cheesesteak and fries as they discuss the highlights of the game.

After their male Mr. Wilson, Charlie and Tom head back to AT&T station then that is they have boarded

the train were on their way back to the train station Mr. Wilson probably looks at his two sons, as well

their faces when satisfaction and memorable day spent at the ballpark.

Windows Voice Recognition

I bit the ballpark

It was not a typical Sunday afternoon at the Wilson household. Mr. Wilson wanted to trade his two sons

Charlie and talk to a beautiful day at the ballpark after finishing another school year.Charlie, a junior at

central box east high school two of the top league baseball game as he was so studious.Charlie’s

younger Brother Tom had been too tough to lose games before, but this time he was very excited to

have a father and Sunday at the ballpark. Mr. Wilson thought it was a great idea to take sap to training

and said drive into the ballpark because is very hard to park at the train station. Esther Wilson, Charlie

and Tom got into the car and drove to the train stationMr. Wilson wanted to trade his two sons to ice

cream cone before the game because it was a very hot and humid day. Mr. Wilson’s favor ice cream

flavors mint chocolate chip, charlie’s favors. Along the way they saw ice cream parlor. better pick on

and tom’s favors code using cream. After ordering their ice cream Mr. Wilson and his two sons Saturday

by picnic table And injure their ice cream cone will talk about who would be the starting pitchers for

today’s game Charlie and Tom we’re very excited to spend time with their father ‘cause during the

school year they really do not have a chance tube on. After finishing their ice cream cone, The Wilsons

got back in the car into new driving to the train station. But is known as a broad street one Charlie

wanted to sit at the back of the train was brother and father wanted to sit on the front. To solve this a

llama, Mr. Wilson flipped a coin. Charlie chose heads book when Leonid tales Tom had won the coin

toss, reading them to send the five of the train On the train there were a lot of phone a few Phillies fans,

and everyone was very happy and anxious to get the ball park early. The reason why everyone on the

train wanted to get the ball park early Was because it was pictured day with the Philadelphia Phillies and

their mascot, the Phillie fanatic. Charlie and Tom had taken if the target class in school this past

semester And wanted to show off their talents. On the way the train had to stop at several stops as it

made its way to ATT station, Which is the train station for all Philadelphia stadiums. Once everyone got

off the train, They read the escalator up to street lawful. As they went through the revolving doors To

their left this all the beautiful AG magnificent citizens bank ballpark. It was only a short walk to the

ballpark and along the way they saw street musicians and street vendors. After getting their ice cream

cone Mr. Wilson Charlie and Tom were very thirsty They decided to teach by a bottled water to

consume as they waited to get Inside the stadium. Once they finish their bottled water it was time to

open the gates to go inside. Mr. Wilson pulled out the three tickets that he bought from STUBHUB the

night before. But before they enter the gate, Charlie wanted to buy a program with a score card to keep

track of the game. This is something that he always enjoyed doing and watching a baseball game and

how. The program was scorecard cost $5.00 and Charlie pulled out a $20.00 bill that he earned for

getting straight A’s Per Charlie it felt so rewarding to work hard at school and be able to spend his

money.As Mr. Wilson, Charlie and Tom were walking into the stadium after hitting their ticket to the

stadium attending, install a full.Charlie and taught we’re very excited wait in line to the Phanatic’s

picture As they got to the front of the wind, where the stadium attendees offered to take Mr. Wilson

Page 42: Voice Recognition and Mobility in the Legal Industry

37

Charlie and Tom picture together with the Phanatic As they continue down the hallway, we saw some of

the flood of Phillies are a table signing autographs They were former and current Phillies there, such as

Mike Schmidt, jimmy Rolland, Greg Lubinski And Dominique brown as walls of the fanPhillies TV

broadcasters, cart the Matt stairs and Jamie Moyer This was a thrill for any phone a few Phillies fan as a

reverie waited patiently in line to get their picture taken to get an auEsther Wilson Charlie and Tom

were able to get everyone’s autograph there have a permit souvenir controlling As they were finishing

up in line we still have about 1:15:00 to go for the first and Mr. Wilson Charlie and Tom were still hungry

and one of the ball park franks hopped on The Wilsons week in and out of the crowds to get to the hot

dog stand Bellevue, Once they got to the hot dog stand everyone had to choose with toppings they

wanted Mr. Wilson chose mustard Russian ideas on this hot dog while his two sons chairs ketchup and

onion Mr. Wilson pay for the hot dogs which include a soft drink and made their way up to their seats

One of the most fun things to do when going to baseball games watching batting practice Visiting team

in New York Mets were on the field taking batting practice of giving flyballs With the binoculars and Mr.

Wilson brought you was able to see all the players up close in another secret that a respect for spaWe’re

about another 20 minutes the New York Mets continue to take batting practice while the filly for when

you take the field After the Mets went back to the dugout the Phillies took a few Mr. Will seek a seat

Dominique brown Carlos release jimmy Rollins and chase Utley waiting behind the batting cage It was a

sunny day about 80° there was a beautiful day for bobby Ravel 30 minutes to fill each practice are filled

in batting swing and it was time to change into regaining Of a change in making glassman minute

preparations before the game on the Jumbotron they were showing highlights the previous weMr.

Wilson Charlie and Tom Bahr very fast EU of the highlights and claps activity was no Italy Phanatic bank

came onto the field loans four-wheeler written to the fan Phanatic and when often as four-wheeler and

watch the pitcher’s mound were two children when they threw out the ceremonial firstBy no plate or

killed a few catchers Carlos Ruiz run the address The two children through the first pitch with varying

stride and the two catchers and add to the catch the balls After the Trojan up children throughout the

first pitch into Phillies catchers went to the pitcher’s mound agreed to children inPublic address

announcer from citizens bank ballpark announced that the phone a few youth choir All the fans in the

ballpark were as their feet as a lot of America Afterward to two teams band is better home plate Agreed

to discuss the rules of the ballpark to home plate umpire Again was about to begin A starting pitchers

were hopefully But the Phillies and Jonathan Niese for the Mets First batter for the mats was Curtis

Granderson

After four pitch it that way and obstructing careers and I’ll Following two hairs PERTAMINA naturally A

salsa bar is a game and so the game quickly became a pitchers’ duel Of fully in the story giving up a few

hits both teams offenses found it difficult to score It wasn’t so the bottom of the seventh inning when

the Phillies finally put one up on the scoreboard Phillies hot on to their one-run we long enough for the

closure Johnson Papelbon the calm in the top of the ninth and close it oPhillies get a win one and often

and how Mr. Wilson has two sons Clap and cheer as they watch the team congratulate each other They

excitedly washed on the Jumbotron recording of the late Harry Kalas and high hopes as a ballpark

crucially plays after how Well as excitement sharing seem to stir up an appetite and Mr. Wilson and his

sons so they decided to celebrate the victory withThey all enjoy their cheesesteak and fries as they

discuss the highlights of the game After their meal Mr. Wilson Charlie and Tom added back to 18 T

station 10 minutes they have boarded the train they were on their way back to the train station Mr.

Wilson pro and looked at his two sons and smiles on their faces when dissatisfaction ruled a span of the

ball

Page 43: Voice Recognition and Mobility in the Legal Industry

38

The Original Script (Edited By Dragon Naturally Speaking)

A Day at the Ballpark

It was not a typical Sunday afternoon in the Wilson household. Mr. Wilson wanted to treat his two sons

Charlie and Tom to a beautiful day at the ballpark after finishing another school year. Charlie, age 16,

and Tom, age 12, always wanted to go to Citizens Bank Ballpark in Philadelphia, PA, but have never

gotten the chance to go. Charlie, a junior at Central Bucks East High School, never got the opportunity to

go to a Philadelphia Phillies baseball game because he was so studious. Charlie's younger brother Tom

had been to Philadelphia Phillies games before, but this time he was very excited to have a father and

son day at Citizens Bank Ballpark. Mr. Wilson thought it was a great idea to take the SEPTA train instead

of driving to the ballpark because it's very hard to park at the stadium and parking is cheaper at the train

station. Mr. Wilson, Charlie and Tom got into the car and drove to the train station. Along the way they

saw an ice cream parlor. Mr. Wilson wanted to treat his two sons to an ice cream cone before the game

because it was a very hot and humid day. Mr. Wilson's favorite ice cream flavor is mint chocolate chip,

Charlie's favorite is butter pecan, and Tom's favorite is cookies and cream. After ordering their ice

cream Mr. Wilson and his sons sat at a nearby picnic table and enjoyed their ice cream cone while

talking about who would be the starting pitchers for today's game. Charlie and Tom were very excited to

spend time with their father because during the school year they really do not have a chance to bond.

After finishing their ice cream cone, the Wilsons got back in the car and continued driving to the train

station. Once they got there, they boarded the Orange Line, which is known as the Broad Street Line.

Charlie wanted to sit in the back of the train, but his brother and father wanted to sit in the front. To

solve this dilemma, Mr. Wilson flipped a coin. Charlie chose heads but the coin landed tails. Tom had

won the coin toss, leading them to sit in the front of the train. On the train there were a lot of

Philadelphia Phillies fans, and everyone was very happy and anxious to get to the ballpark early. The

reason why everybody on the train wanted to get to the ballpark early was because it was picture day

with the Philadelphia Phillies and their mascot, the Phillie Phanatic. Charlie and Tom had taken a

photography class in school this past semester and wanted to show off their talents. On the way the

train had to stop at several stops as it made its way to AT&T station, which is the train station for all

Philadelphia stadiums. Once everyone got off the train, they rode the escalator up to street level. As

they went through the revolving doors, to their left they saw the beautiful and magnificent Citizens Bank

Ballpark. It was only a short walk to the ballpark and along the way they saw street musicians and street

vendors. After eating their ice cream cone, Mr. Wilson, Charlie, Tom were very thirsty. They decided to

each buy a bottled water to consume as they waited to get inside the stadium. Once they finish their

bottled water, it was time to open the gates and go inside. Mr. Wilson pulled out the three tickets that

he bought from STUBHUB the night before and gave one ticket to each of his sons. But before they

entered the gate, Charlie wanted to buy a program with a scorecard to keep track of the game. This is

something that he always enjoyed doing when watching a baseball game at home. The program with

scorecard cost five dollars and Charlie pulled out a $20 bill that he earned for getting straight A's on his

last report card. For Charlie it felt so rewarding to work hard at school and be able to spend his money

and his time with people he enjoyed being around. As Mr. Wilson, Charlie and Tom were walking into

the stadium after handing their ticket to the stadium attendee, they saw the Phillie Phanatic. Charlie and

Tom were very excited and waited in line to take the Phanatic's picture. As they got to the front of the

line, one of the stadium attendees offered to take Mr. Wilson, Charlie, Tom's picture together with the

Phanatic. As they continued down the hallway, they saw some of the Philadelphia Phillies at a table

signing autographs. There were former and current Phillies there, such as Mike Schmidt, Jimmy Rollins,

Page 44: Voice Recognition and Mobility in the Legal Industry

39

Greg Lubinski, and Dominic Brown, as well as the Philadelphia Phillies TV broadcasters Tom McCarthy,

Matt Stairs, and Jamie Moyer. This was a thrill for any Philadelphia Phillies fan as everybody waited

patiently in line to get their picture taken and to get an autograph. Mr. Wilson, Charlie and Tom were

able to get everyone's autograph there and had a permanent souvenir to treasure. As they were

finishing up in line they still had about an hour and 15 minutes to go before the first pitch. Mr. Wilson,

Charlie and Tom were still hungry and wanted a Ballpark Franks hot dog. The Wilsons weaved in and out

of the crowd to get to the hot dog stand located at the above concourse. Once they got to the hot dog

stand, everyone had to choose what toppings they wanted. Mr. Wilson chose mustard, relish and onions

on his hot dog while his two sons chose ketchup and onions. Mr. Wilson paid for the hot dogs which

included a soft drink and they made their way up to their seats. One of the most fun things to do when

going to a baseball game is watching batting practice. The visiting team, the New York Mets, were on the

field taking batting practice and fielding flyballs. With the binoculars that Mr. Wilson brought, he was

able to see all the players up close, even though their seats were 30 rows back from first base. For about

another 20 minutes, the New York Mets continue to take batting practice while the Phillies were waiting

to take the field. After the Mets went back to their dugout, the Phillies took the field. Mr. Wilson could

see Dominic Brown, Carlos Ruiz, Jimmy Rollins and Chase Utley waiting behind the batting cage. It was a

sunny day; about 80°, and was a beautiful day for a ballgame. For about 30 minutes, the Phillies

practiced their fielding and batting swings and it was time to change into their game uniforms. While

both teams were making last preparations before the game, on the Jumbotron they were showing

highlights from the previous week in Major League Baseball. Mr. Wilson, Charlie and Tom were very

fascinated with the highlights and clapped after the video was over. The Phillie Phanatic then came onto

the field on his four-wheeler and drove around the entire infield waving to the fans. The Phanatic then

went off of his four-wheeler and walked to the pitcher's mound where two children were waiting to

throw out the ceremonial first pitch. Behind home plate were Philadelphia Phillies catchers Carlos Ruiz

and Wil Nieves. The two children threw the first pitch with varying strength as the two catchers

attempted to catch the balls. After the two children threw out the pitch, the two Phillies catchers went

to the pitcher's mound to greet the children and the Phanatic. The public address announcer from

Citizens Bank Ballpark announced that the Philadelphia Youth Choir would be singing the national

anthem. All the fans in the ballpark rose to their feet as they honored America. Afterwards, the two

teams' managers met at home plate, shook hands and exchanged batting lineups. They briefly discussed

the rules of the ballpark with the home plate umpire. The game was about to begin. The starting

pitchers were Cliff Lee for the Phillies and Jonathon Niese for the Mets. The first batter for the Mets was

Curtis Granderson.

After a four pitch at bat Lee ended up striking Granderson out. The following two hitters proved to be

no match for Lee, as the bottom of the first quickly approached. Niese also brought his A-game, and so

the game quickly became a pitcher's dual. With both Lee and Niese only giving up a few hits, both

teams' offenses found it difficult to score. It wasn't until the bottom of the seventh inning when the

Phillies finally put one up on the scoreboard, thanks to Carlos Ruiz's sac fly that brought home Marlon

Byrd. The Phillies hung on to their one-run lead long enough for their closer Jonathan Papelbon to come

in the top of the ninth and close it out with two strikeouts and a groundout. The Phillies get a one to

nothing home victory. Mr. Wilson and his two sons clap and cheer as they watch the team congratulate

each other on their win. They excitedly watch on the Jumbotron a recording of the late Harry Kalas

singing "High Hopes," as the ballpark traditionally plays after a home victory. All of this excitement and

cheering seemed to stir up an appetite in Mr. Wilson and his sons; so they decide to celebrate the

victory with a satisfying meal from Philadelphia's very own Chickie's and Pete's. They all enjoy their

cheesesteak and fries as they discuss the highlights of the game. After their meal Mr. Wilson, Charlie

and Tom head back to AT&T station. Within minutes they have boarded the train and were on their way

Page 45: Voice Recognition and Mobility in the Legal Industry

40

back to the train station. Mr. Wilson proudly looks at his two sons, as the smiles on their faces display

the satisfaction of a memorable day spent at the ballpark

Page 46: Voice Recognition and Mobility in the Legal Industry

41

Appendix 2: Initial Case Briefing

During this exercise, our team was given a normal Initial Case Briefing Template that legal assistants and

attorneys work with on a daily basis. This exercise was to compare the length of time it took from start

to finish to compose and edit the template to its original state and compare our statistics to other

published blogs on words per minute (wpm) versus Voice Recognition Software WPM. Below is the

template which we worked with:

Initial Briefing Legal Document Dictation from a Template December 26, 2013

VIA EMAIL

Ms. Adjuster

Adjuster Company

P.O.Box 420

Dallas, TX 75265

Re: Claimant v. Company

Date of Loss: 5/21/2009

Insurer Claim Number: 7506231–23

Our File Number: 006732

Dear Ms. Adjuster:

please adjust this letter as our analysis of the file documents sent to us in the above captioned

matter.

BACKGROUND

Claimant was born on September 24, 1963 and is 50 years old. He was hired by Company, Inc.,

on January 4th 2011 and worked as a control tech. He sustained an injury to his left knee on June 3, 2011

when he was bending down to pick up a glue board and heard a “pop”.

BUREAU DOCUMENTS

The Bureau documents consists of a First Report of Injury and a notice of Temporary

Compensation Payable (TNCP). The TNCP was issued on June 20, 2011 acknowledging the knee strain.

Although the TNCP is not specify the left knee, all the treatment records indicate that the injury was the

claimants left knee. The average weekly wage is $700.42 and the weekly compensation rate is $550.28.

There is no indication that the TNCP was stopped. The claimant has never returned to work, the

payment of his benefits continues at the present time.

MEDICAL RECORDS

On August, 2011, claimant was treated in the emergency room of the Holy Redeemer Hospital.

The history given by claimant states that he squatted down and hurt his left knee at work. The physical

examination was essentially normal, and he was discharged to follow up with his primary care physician.

The diagnostics was a sprained knee.

On August 10, 2011 claimant was seen by Dr. Trey Anastasio and was referred to the Rothman

Institute. Dr. Michael Gordon ordered an MRI, which was performed on August 18, 2011 and was

interpreted by Jon Fishman as showing a medical meniscus tear and edema in the medical tibial plateau.

A steroid injection administered. Dr. Page McConnel at the Rothman Institute saw the claimant on

September 14, 2011. His review of the MRI showed the medical meniscus tear and, in addition, a

probable bone marrow lesion porch/insufficiency fracture of the medical tibial plateau. Dr. McConnel

recommended and perform surgery on September 22, 2011 at Nazareth Hospital. The arthroscopic

procedure included repair of the Medio tibial plateau insufficiency fracture but drilling and grafting as

Page 47: Voice Recognition and Mobility in the Legal Industry

42

well as a partial medical meniscus meniscectomy. Claimant was sent home following the procedure but

was readmitted for “excruciating pain” the same day. The claimant continue to follow up with Dr.

Robert Mould and Dr. McConnel at the Rothman Institute for the post – operative pain, and on October

10, 2011 Dr. Mould detected a lump in the claimants left thigh. On November 7, 2011 Dr. Mould stated

that the problem in the claimants left thigh was not related to the work injury. In this time frame,

claimant was treated with medication of physical therapy. As of December 20, 2011, Dr. Mould

completed a form stating that the claimant was temporarily totally disabled from employment activities.

INDEPENDENT MEDICAL EXAMINATIONS

Claimant had an IME by Dr. Chris Kuroda, a board certified orthopedic surgeon, on January 26,

2012. Following his review of the medical records and examination, Dr. Kuroda concluded that the

claimant had not reached MMI was not fully recovered. He recommended additional diagnostic testing.

Dr. Corona believed that claimant was capable of working in the sedentary classification. Dr. Corona

concluded that ongoing physical therapy was not reasonable or necessary.

ANALYSIS

The available medical records indicate that the description of injury will need to be amended to

include a torn medial meniscus of the left knee as well as the RSD/CRPS diagnosed by Dr. Kuroda’s most

recent IME.

The available medical record support requesting an IRE at this time which, if successful, will be

the first step in reducing future wage loss liability and positioning the case for possible settlement.

There are no petitions which can be filed at this time.

RECOMMENDATIONS

1) Request an IRE

2) Send LIBC reporting forms.

3) Conduct social media investigation and activities check to see if we can find anything really

before we pay big money on this case

4) Ascertain from claimant’s counsel any interest in settlement and claimants Medicare

5) if settlement interest exists, obtain a demand and value the case for settlement. If no interest

in settlement exists, refer the case for a vocational evaluation and Labor Market Survey/Earning Power

Assessment.

Thank you again for the referral of this case. We look forward to working with you to bring this

matter to a favorable conclusion. After you have the opportunity to review this analysis, please contact

me to discuss recommendations given above.

Reflections after Dictating Initial Briefing Analysis The above Initial Briefing Analysis took about an hour to complete and edit as dictated to the document

using a standard format. Dragon Naturally Speaking continues to evolve with the user’s profile

increasing its accuracy as words are spoken. There is a learning curve for beginners who are using

Dragon Naturally Speaking which includes making voice corrections within a document. For a beginner, a

user may think that every punctuation mark needs to be dictated such as the comma in Philadelphia, PA.

Dragon Naturally Speaking Software can recognize when a city and state or a date is dictated so the

comma can be omitted from the dictation and still appear on the document.